Smalot\PdfParser\Document::buildDetails()

Complexity

Conditions	10
Paths	63

Size

Total Lines	68
Code Lines	29

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	27
CRAP Score	10.0045

Importance

Changes	1
Bugs	1	Features	0

<?php
    protected function buildDetails()
    {
        // Build details array.
        $details = [];

        // Extract document info
        if ($this->trailer->has('Info')) {
            /** @var PDFObject $info */
            $info = $this->trailer->get('Info');
            // This could be an ElementMissing object, so we need to check for
            // the getHeader method first.
            if (null !== $info && method_exists($info, 'getHeader')) {
                $details = $info->getHeader()->getDetails();
            }
        }

        // Retrieve the page count
        try {
            $pages = $this->getPages();
            $details['Pages'] = \count($pages);
        } catch (\Exception $e) {
            $details['Pages'] = 0;
        }

        // Decode and repair encoded document properties
        foreach ($details as $key => $value) {
            if (\is_string($value)) {
                // If the string is already UTF-8 encoded, that means we only
                // need to repair Adobe's ham-fisted insertion of line-feeds
                // every ~127 characters, which doesn't seem to be multi-byte
                // safe
                if (mb_check_encoding($value, 'UTF-8')) {
                    // Remove literal backslash + line-feed "\\r"
                    $value = str_replace("\x5c\x0d", '', $value);

                    // Remove backslash plus bytes written into high part of
                    // multibyte unicode character
                    while (preg_match("/\x5c\x5c\xe0([\xb4-\xb8])(.)/", $value, $match)) {
                        $diff = (\ord($match[1]) - 182) * 64;
                        $newbyte = PDFDocEncoding::convertPDFDoc2UTF8(\chr(\ord($match[2]) + $diff));
                        $value = preg_replace("/\x5c\x5c\xe0".$match[1].$match[2].'/', $newbyte, $value);
                    }

                    // Remove bytes written into low part of multibyte unicode
                    // character
                    while (preg_match("/(.)\x9c\xe0([\xb3-\xb7])/", $value, $match)) {
                        $diff = \ord($match[2]) - 181;
                        $newbyte = \chr(\ord($match[1]) + $diff);
                        $value = preg_replace('/'.$match[1]."\x9c\xe0".$match[2].'/', $newbyte, $value);
                    }

                    // Remove this byte string that Adobe occasionally adds
                    // between two single byte characters in a unicode string
                    $value = str_replace("\xe5\xb0\x8d", '', $value);

                    $details[$key] = $value;
                } else {
                    // If the string is just PDFDocEncoding, remove any line-feeds
                    // and decode the whole thing.
                    $value = str_replace("\\\r", '', $value);
                    $details[$key] = PDFDocEncoding::convertPDFDoc2UTF8($value);
                }
            }
        }

        $details = array_merge($details, $this->metadata);

        $this->details = $details;
    }